Subsurface characterization & foundations

Before any reserves or production figure means anything, you have to understand the rock and the data that describes it. This collection is the foundation: the geology and geophysics that frame a prospect, the petrophysics that turns logs into rock properties and pay, and the data-science layer that increasingly sits on top of all of it.

In this collection

Regional geology for non-geologists — the basin-and-prospect framing every subsurface decision rests on.

Seismic for non-geophysicists — reading the seismic picture without the heavy mathematics.

Rock typing, reimagined — flow units from logs and core, with a modern Python and analytics workflow.

Low-resistivity & low-contrast pay — the pay that a quick-look resistivity scan misses — and how to quantify it.

AI & machine learning in upstream oil & gas — where ML genuinely helps in the subsurface, and the pitfalls that sink naive models.

Each article stands on its own, but together they build a single line of reasoning. Read them in order for a structured tour, or jump to the one that matches the problem in front of you.

Frequently asked questions

What is formation evaluation?

Formation evaluation determines reservoir rock and fluid properties — porosity, permeability, water saturation and net pay — from logs, core and seismic, to establish what is present and whether it will produce.

What is rock typing?

Rock typing groups reservoir rock into units with similar pore geometry and flow behaviour (flow units), linking core measurements to logs so permeability and saturation can be predicted across the field.

How is machine learning used in the subsurface?

For log prediction, facies classification and property inference — most reliably when it is physics-informed and validated against held-out data to avoid leakage, tiny-data overfitting and extrapolation errors.

Why is seismic important for non-geophysicists?

Seismic provides the structural and stratigraphic picture that frames where reservoir and traps are likely to be, so even non-specialists benefit from reading its key features when making subsurface decisions.